Comments on: The C++ preprocessor – undefined tokens http://nealabq.com/blog/2008/08/13/the-c-preprocessor-undefined-tokens/ ... dodging grues in the dark Tue, 27 Dec 2011 18:28:32 +0000 hourly 1 http://wordpress.org/?v=3.0 By: Neal http://nealabq.com/blog/2008/08/13/the-c-preprocessor-undefined-tokens/comment-page-1/#comment-65 Neal Tue, 30 Dec 2008 18:13:52 +0000 http://nealabq.com/blog/?p=236#comment-65 Yeah, seems like a bad idea to me too. I suspect it came about because some C compilers back in the loose-standards day were doing this. Dunno, maybe they wanted to avoid short-circuited evaluation of the #if test. But I'm pretty sure the standards guys thought and fought about it, and had good reasons at the time. Still, small consolation after you've wasted hours hunting down a bug. Yeah, seems like a bad idea to me too. I suspect it came about because some C compilers back in the loose-standards day were doing this. Dunno, maybe they wanted to avoid short-circuited evaluation of the #if test.

But I’m pretty sure the standards guys thought and fought about it, and had good reasons at the time.

Still, small consolation after you’ve wasted hours hunting down a bug.

]]> By: yonil http://nealabq.com/blog/2008/08/13/the-c-preprocessor-undefined-tokens/comment-page-1/#comment-64 yonil Tue, 30 Dec 2008 08:20:11 +0000 http://nealabq.com/blog/?p=236#comment-64 after running into this issue myself, and spending a couple hours hunting a bug related to a side effect of this NOT_DEFINED issue, I'm bewildered. what was the standards commitee thinking when they allowed the preprocessor to substitute 0 for undefined tokens? it should omit a level 1 warning at minimum. after running into this issue myself, and spending a couple hours hunting a bug related to a side effect of this NOT_DEFINED issue, I’m bewildered. what was the standards commitee thinking when they allowed the preprocessor to substitute 0 for undefined tokens? it should omit a level 1 warning at minimum.

]]>